APPENDIX 39
Memorandum submitted by Professor Gary
Craig, Professor of Social Justice, University of Hull
WHAT DOES IT ALL MEAN?
SOCIAL POLICY AND THE 2002 RESEARCH ASSESSMENT
EXERCISE
Before the RAE 2001 results were announced,
there were already anxious questions being asked about the strength
of the discipline which it was felt might be addressed by the
exercise. From data supplied by HEFCE to the SPA in the autumn,
it was already clear that there had been considerable change in
the list of those submitting to social policy, reinforcing the
picture which had been emerging over the past two years. Nine
HEIs which submitted to the 1996 exercise had dropped out of the
reckoning; most of these were new universities with a grade of
two or less in 1996 and vice-chancellors had obviously decided
that the game wasn't worth the candle. Some of them were institutions
where undergraduate social policy teaching has been terminated
in the past few years and some had seemingly submitted social
policy staff elsewhere, notably to sociology, which also suggested
a downgrading of social policy within the HEIs concerned. A total
of about 87 staff, about 13 per cent of all FTE research-active
social policy staff submitted in 1996 were "lost" in
2001 as a result of these disappearances (these staff may not
have left the discipline but had not been submitted to the RAE).
All of this was not good news to an area of study which has appeared
in recent years to be losing ground and losing its distinct identity.
The professional community entertained hopes that the RAE would
provide a welcome boost to its image.
Other early indications were more positive.
In 1996 there were 44 submissions to social policy. In 2001, there
were 47 submissions to social policy; not much change apparently
there then. However, even with the nine HEIs from 1996 no longer
present, a substantial new number submitted to social policy in
2001 which had not done so in 1996 and a total of almost 850 staff
were submitted, more than 30 per cent up on 1996. This appeared
to suggest a healthy growth in the discipline, contradicting the
worrying news of course closures and difficulties in recruiting
undergraduates. However, closer inspection also suggested that
even these indications were not as positive as seemed at first
sight.
New submissions in 2001 ie those from HEIs which
did not submit to the 1996 social policy panel, could be grouped
into four clusters as follows:
(a) very new to the RAE field: Bolton Institute,
Chichester University College (20 staff in all);
(b) new to the RAE social policy field: City,
Plymouth, Westminster, Paisley: all small staff teams (five-nine),
totalling 30 staff.
(c) HEIs which appear to have done a straight
swap from social work or some other panel in 1996 to social policy
in 2001: Birmingham had five in social work and eight in social
policy in 1996 but submitted 16 social policy and social work
to the social policy panel in 2001; Bradford (11 in both 1996social
workand 2001); Liverpool (14 in 1996social work,
18 in 2001); Nottingham (24 in sociology in 1996, 18 in 2001);
Nottingham Trent (seven in sociology in 1996, 18 in 2001): these
shifts may have reflected changed staff priorities or tactical
moves or a combination of both. Overall, these shifts to social
policy appeared to bring 73 new staff into the social policy field.
(d) Fairly seismic shifts:
(1) Birmingham: Birmingham put two separate submissions
to the social policy panel, one of which, a new public policy
submission, has 51 staff, presumably gathered up from various
other places not submitted to social policy previously. This submission
sits alongside a separate social policy submission from Birmingham,
an apparent case of having your cake and eating it.
(2) Leeds submitted 22 to sociology in 1996 (grade
four), 48 to social policy in 2001.
(3) Southampton submitted 17 to sociology in
1996 (grade four), 26 to social policy in 2001.
(4) University of Wales, Swansea submitted 15
to social work in 1996 and 10 in 2001 but none to social policy
(or sociology) in 1996, and 15 to social policy in 2001.
These reconfigurations may also reflect
a combination of tactical moves, the outcome of arguments within
HEIs (and particularly those which faced the difficulty of dealing
with staff in joint departments covering two or more separate
disciplines), and shifts in staffing profiles but are particularly
marked here by the numbers of staff involved.
The overall effect of these "new"
submissions was about 260 new staff recorded as social policy
academics in the RAE. The average number of staff per submission
was 18 (up from about 15 in 1996). In summary, then, the picture
was of more submissions, bigger submissions; seemingly good news.
However, in the parallel context of social policy's
well-rehearsed difficulties as a discipline, it was hard to avoid
asking what proportion of those staff new to the social policy
submissions were recognisably social policy academics. In the
1996 RAE, whilst the social policy community were supportive of
the attempts of the panel to accept a wide-ranging understanding
of research quality, mutterings were heard from colleagues in
sociology that they had had a rough ride and it would have been
perfectly understandable if some HEIs had rebranded their groupingsas
opposed to reorienting their research prioritiesin order
to submit to social policy. These questions may only be answered
in the medium term.
When the results were announced in mid-December,
more fundamental questions yet have been posed by social policy's
"performance". Of the 47 submissions, 16 (about one
third) attained the same grade in 2001 as in 1996. 13 went up
one grade, two (North London and Sunderland) went up two. Two
(Open and Bangor) dropped a grade. 14 of the 47 had no social
policy grade in 1996 and therefore, in reality, for half of those
33 HEIs submitting in both 1996 and 2001, the picture was one
of standstill. With the exception of Plymouth, none of those new
(second submission only) or very new (first submission) to the
social policy panel gained a grade higher than 3a, as one might
have expected. Amongst the five HEIs which appear to have done
a straight swap from social work or some other panel in 1996 to
social policy in 2001, four achieved a grade 4, one had a 3a.
And of those HEIs newly submitting large numbers of staff to social
policy, one gained a 3a, one a grade 4 and two gained a 5; a total
of 140 staff, or about one in six of the total submitted to social
policy were involved in these last four submissions.
Of the submissions for 1996 and 2001, the distribution
of grades was as follows:
No achieving grade
| 1996 | 2001
|
5* | 1 (1)
| 2 (2) |
5 | 4 (3)
| 8 (6) |
4 | 13 (13)
| 16 (10) |
3a | 6 (5)
| 15 (11) |
3b | 12 (10)
| 4 (4) |
2 | 6 (2)
| 2 (0) |
1 | 2 (0)
| |
This table offers the basis for two types of comparison.
The figures in brackets in each column show the spread of grades
obtained by those institutions submitting to both 1996 and 2001
exercises and probably gives the clearest sense of how the core
social policy communityin one definitionhas developed
over the five year period measured by RAE indicators. This does
show a very modest upward grade drift indeed, particularly between
4 and 5 grades, but there is also a clear bifurcation: half in
each exercise were 4 and above, half 3a and below. In 2001, there
were some notable casualties, the drop at the Open University
of one grade and the failure of Brighton to improve on its 1996
grade, both causing eyebrows to be raised in considerable surprise
amongst the community. This overall picture is in any case muddied
considerably by the 2001 "imports"; the overall picture,
taking all submissions into account suggests a much strongerif
substantially misleadingpicture of upward grade drift.
HEFCE has now announced that it will protect 5* departments,
offer less than in 1996 to 5 and 4 rated departments, little to
3a departments and nothing to 3b departments. HEFCE has also announced
it is not to implement any decisions on funding for a year, presumably
as a tactic to bid up in negotiations with government for the
higher education settlement. This is where the profile of social
policy vis-á-vis other cognate disciplinesand indeed
all other disciplinesgives particular cause for alarm and
raises yet further questions about the 2001 results. What do they
mean to the community and, of equal long-term significance, how
will they be interpreted by those responsible for funding research?
Fifty-five per cent of all staff submitted to RAE across
the whole exercise were in 5 or 5* rated units. In social policy,
however, only 35 per cent of all staff submitted (including submissions
"new" to social policy as described earlier) were in
5 or 5* departments (up from 21 per cent in 1996), compared with
38 per cent for social work (21 per cent in 1996), and, ironically,
given the apparent flight of a substantial number of submissions
from sociology to social policy, 52 per cent also in sociology
(24 per cent). In law, an apparently astonishing 84 per cent of
staff submitted are to be found in 5 or 5* departments but this
appears to be because the law panel took a narrow definition of
qualityparticular kinds of cited research output. Only
six other units of assessment of the total of 69 registered a
smaller proportion of staff than social policy in 5 and 5* departments.
Does this really reflect a poor performance by the social
policy community?; and what will be the impact on funding for
social policy research through the RAE? Any interpretation of
these results can only be speculative at this moment. This contribution
is made by someone who, although he read several RAE submissions
in draft from HEIs other than his own, has no privileged information
and is, along with colleagues elsewhere, searching for a positive
interpretation of the results which appear to fly in the face
of his own experience of the research profile of social policy.
SPA hopes that other members might like to add to the debate in
succeeding issues to SPAN to give shape to a policy campaign for
the Executive.
Explanations offered to the present author to date include
the following and, in light of detailed engagement with the panel
members, with members of other panels and with HEFCE's own views,
more may present themselves as plausible explanations as time
goes by.
(i) the quality of social policy research really is poor
in the sense that it has fallen well behind virtually every other
discipline measured by improvements in RAE ratings. It is difficult
to imagine many within the social policy community adhering to
this view but some self-reflection is clearly appropriate;
(ii) other panels have adopted a much more strongly protective
approach to their discipline by adopting means of working, and
judgments of quality, which led to substantial hikes in RAE ratings
across the piece. Without knowing how all other panels worked,
it is difficult to address how this might have happened but clearly
the social policy panel will want to be as open as possible to
their own constituency about their ways of working, how they might
have differed, if at all, from that of the 1996 social policy
panel, for example in adopting the approach of being a reading
panel (as opposed to the "reading-across" approach of
some panels which took publication in particular outputs as a
de facto indicator of quality), and any insights they may have
on ways in which other cognate panels may have taken differing
approaches;
(iii) the problem of the common definition of 5 and 5*
quality, which involves a high international provenance. As my
report on the 1996 exercise (Craig 1996[10])
argued, one of social policy's great strengths is its local policy
relevance and there can be little doubt that this has grown in
a period when a huge range of government initiatives have needed
research and evaluation and where government has pressed the urgency
of its social policy agenda on the community. Quite apart from
this contemporary agenda, much social policy research is necessarily
national because of its empirical, policy and theoretical base.
This was used as an argument in my report against the further
concentration of resources into a relatively few HEIs but in a
context where there have apparently been remarkable improvements
in other disciplines judged by this international standard, members
of the social policy panel may have felt structurally constrained
to judge research quality by inappropriate criteria. What might
would lead to de facto would be the penalising of HEIs
with a relatively "weak" international/comparative research
record which the use of high profile international referees would
simply accentuate.
On the face of it, and without further information, the 2001
RAE results may appear to have damaged the profile of social policy
and a sustained campaign by the Social Policy Association acting
in concert, one hopes, with members of the RAE panel, will be
needed to place an appropriate interpretation on the results and
defend our corner in the forthcoming discussions on research funding.
The alternative may be that the overall pot of money to social
policy will be inadequateeven by the generally parsimonious
constraints within which HEFCE has to workand its distribution
within the social policy community even more inequitable than
hitherto. Few in the social policy community would claim that
a piece of high profile comparative research is necessarily better
quality than a report for the Department of Health or a local
study funded by a charitable body or a consortium of local authorities.
Yet that continues to be the implication of the general measures
used by HEFCE and social policy may well suffer as a result. The
trend, re-emphasised by the 2001 results, towards two tiers of
social policy departments, in which only the top tier is seen
as undertaking research worthy of HEFCE funding, may suit some
of those in the top tier but would do nothing for the long-term
sustainability of the discipline as a whole. Social policy has
always argued that it needs both high quality and locally-relevant
research; but it also needs a critical mass, in terms of its presence
in a wide range of HEIs, if it is to be sustainable in terms of
attracting resources. The alternativeas appears to be happening
already to some degreeis that social policy will become
submerged within other departments or disciplines.
There has been a growing feeling for some time that many
aspects of the RAE are indefensible, not least that it allows
substantial game-playing (for example as to which panel to submit
to, how many staff to enter or not and so on); that it allows
for no appeal save on procedural grounds, leaving substantially
aggrieved departments to live with the consequences of unchallengable
judgments for five years; that it does not encourage risk-taking
and developmental work or indeed the development of new disciplinary
areas; that it reinforces the position of the haves against the
have-nots; and that it is inappropriate to use a single set of
criteria by which to judge the quality of research in more than
sixty different disciplinary areas with different histories, traditions
and culture. The 2001 outcomes appear to have done little to address
these fears and the Social Policy Association now has a substantial
job of work to do to defend the interests of all of its members.
The 2002 RAE raises questions about the measures used within
the exercise, and most of all the usefulness of a single set of
performance indicators used uniformly across all disciplines,
which it is hoped the Select Committee may address in its discussions.
24 January 2002
10
Craig, G., (1996), Quality First?: the assessment of quality
in social policy research, London: Social Policy Association. Back
|